POWER GRID INFRASTRUCTURAL RESILIENCE AGAINST EXTREME WEATHER EVENTS CONSIDERING ENERGY JUSTICE

Abstract

This dissertation presents a comprehensive research effort aimed at improving the resilience and equity when implementing design and planning of power distribution systems under increasing climate hazards. The work introduces a series of interconnected methodologies for vulnerability assessment, infrastructure optimization, and rural electrification, offering practical tools for resilient and just energy system planning.

It begins by critically examining how resilience is defined and measured in power systems. The review highlights the absence of standardization across definitions and metrics, exposing key gaps that hinder consistent and equitable resilience planning. To address this, a new vulnerability metric is developed for overhead power lines by integrating meteorological and geographical factors. Using principal component analysis and Pareto ranking, the proposed method identifies high-risk overhead line segments, supporting targeted and proactive mitigation strategies.

To deepen the analysis, a statistical framework is then proposed that quantifies how environmental and land-based variables may influence outage severity. This approach incorporates weather data, terrain characteristics, and wildfire risk, applying techniques such as multiple regression, LASSO regression, and interaction modeling. The results reveal the most impactful drivers of outages and provide data-driven guidance for resilience investment decisions.

Building on these insights, a hazard-aware multi-objective optimization model is introduced to guide infrastructure reinforcement. The model balances capital investment, outage reduction, and equity considerations by incorporating spatial social vulnerability into the decision-making process. Applied to a distribution system in Greeley, Colorado, the model demonstrates how targeted line hardening strategies can enhance both technical resilience and fairness in resource allocation. In parallel, the research addresses the challenge of rural electrification. A optimization model is used to identify candidate sites for microgrid deployment and/or grid extension by combining technical, economic, environmental, and social criteria. These candidate locations feed into a Chebyshev goal programming model that evaluates trade-offs between cost, emissions, energy access, and social equity. The approach is applied to the Shiprock region of the Navajo Reservation, using empirical grid data and synthetic demand profiles. The results show that combining microgrids and grid extension can efficiently serve remote communities while prioritizing those with higher vulnerability and energy needs.

Energy justice is a central theme throughout this thesis, with a focus on both recognition and distributional justice. The models developed in this work account for social vulnerabilities of communities against lack of access to electricity and aim to ensure that resilience improvements are fairly distributed across the community, avoiding disproportionate negative impacts on any sections of the community. This helps align infrastructure investment with community needs and supports a more just and inclusive energy transition.

Collectively, the methodologies presented in this dissertation form a unified framework for advancing resilient and equitable distribution system planning.